Current methods integrating variant functional annotation scores have limited capacity to improve the power of genome-wide association studies
Bibliographic record
Abstract
Abstract Functional annotations have the potential to increase the power of genome-wide association studies (GWAS) by prioritizing variants according to their biological function. Focusing on variant-specific annotation meta-scores including CADD (Kircher et al., 2014) and Eigen (Ionita-laza et al., 2016), we broadly examined GWAS summary statistics of 1,132 traits from the UK Biobank (Sudlow et al., 2015) using the weighted p-value approach (Genovese et al., 2006) and stratified false discovery control (sFDR) method (Sun et al., 2006). These 1,132 traits were rated by Benjamin Neale’s lab from the Broad Institute as having medium to high confidence for their heritability estimates. Averaged across the 1,132 UK Biobank traits, sFDR was more robust to uninformative meta-scores, but the weighted p-value method identified more variants using CADD or Eigen, based on performance measures that included type I error control, recall, precision, and relative efficiency. Our application results were consistent with those from an extensive simulation study using three different designs, including leveraging the real genetic data combined with simulated genomic data and vice versa. We also considered the recent FINDOR method (Kichaev et al., 2019), which leverages a set of individual 75 functional annotations into GWAS. An earlier application of FINDOR to 27 traits selected from the z7 category (SNP-heritability p-value < 1.27 × 10 −12 by Nealelab) detected 13%-20% additional genome-wide significant loci as compared to the standard annotation-free GWAS, which we confirmed. Moreover, across all 438 traits in the z7 category, 46,631 out 59,764 (80%) significant loci discovered are common across the three data-integration methods. However, across all the 1,132 UK Biobank traits examined, the median [Q1,Q3] of the total numbers of new, genome-wide significant independent loci were 0 [0, 3] by FINDOR, 0 [0, 2] by weighted p-value, and 0 [0, 0] by sFDR. Notably, 162 traits (89%) in the nonsig trait category (SNP-heritability p-value > 0.05, “likely reflecting limited statistical power rather than a true lack of heritability” by Nealelab) had no new discoveries after data-integration by any of the three methods. Our findings suggest that more informative scores or new data integration methods are warranted to further improve the power of GWAS by leveraging the variant functional annotations.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.129 | 0.322 |
| Meta-epidemiology (narrow) | 0.004 | 0.002 |
| Meta-epidemiology (broad) | 0.005 | 0.007 |
| Bibliometrics | 0.009 | 0.014 |
| Science and technology studies | 0.002 | 0.004 |
| Scholarly communication | 0.008 | 0.007 |
| Open science | 0.005 | 0.006 |
| Research integrity | 0.003 | 0.006 |
| Insufficient payload (model declined to judge) | 0.007 | 0.003 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".